RELATED INVENTION

This invention is related to copending patent application, Serial No. 07/961,157, filed October 15, 1992 and entitled POLYMERIC PARTICLES HAVING A BIODEGRADABLE GELATIN OR AMINODEXTRAN COATING AND PROCESS FOR MAKING SAME.

TECHNICAL FIELD

This invention relates generally to polymeric colloidal particles having a short, low molecular weight diamine spacer or a polymeric crosslin ed aminodextran coating that is functionalized to bind a pendent protein such as antibody, which particles are freeze-dried and reconstituted with an aqueous media, to the thermal stability of such particles, to methods of making such particles, and to the use of such particles in biological assays.

BACKGROUND ART

A sensitive measure of polymer structure on colloidal particles is light scattering, related to the size, shape, and refractive index of the scattering object. The modulating effect of a particle coating on light scattering intensity can be significant if the polymeric structures approach the size of the uncoated particles. The Coulter VCS instrument, as described in International Patent Application Publication WO90/13013, to Russell et al., and in PCT application PCT/US91/08590, provides a histogram of radiofreguency (rf) versus median angle light scattering (LS) for lysed and quenched whole blood samples, wherein lymphocytes (L) , monocytes (M) , and granulocytes (G) are separated into non-overlapping regions. Subpopulations of lymphocytes can be enumerated by shifting targeted sub-cell population to a different unoccupied part of the histogram. Monoclonal antibody-

coated polystyrene beads mixed with lysed and quenched whole blood (EDTA anticoagulated) may be used to shift the targeted sub-cell population. The magnitude and direction of this shift will depend on the size, shape, and refractive index of the beads and their relative effect on light scattering from the targeted cells.

Non-targeted cells will also have collisional interactions with these beads. For example, uncoated 2 micron diameter polystyrene beads mixed with lysed and quenched whole blood changed the distributions of lymphocytes, monocytes, and granulocytes cells in the radiofrequency versus median angle light scattering histogram so that each was broadened slightly along the LS axis. Thus, coated beads will have varying broadening effects on targeted and the non-targeted white blood cell populations along the LS axis of the histogram, depending on the relative stickiness of collisions between beads and cells.

Pending patent application Serial No. 07/961,157, discloses the use of aminodextran (T-2M) , referred to as 5X-Amdex, to coat aldehyde/sulfate polystyrene latex beads. In this antibody-spacer-bead system, the spacer is coupled to the aldehyde groups on the polystyrene bead and the antibody is then conjugated to the unreacted amino groups on the surface of the 5X-Amdex-polystyrene bead. T4 and T8 antibody are conjugated to these coated beads, which are then used on the Coulter VCS instrument to give CD4 and CD8 percentages in total lymphocyte populations in whole blood. Broadening in the light scatter histogram of unshifted and non-targeted lymphocytes, monocytes, and granulocytes in the rf versus LS histograms was apparent in the use of T4, T8-5X-Amdex-bead.

Broadening along the light scattering axis of histogram causes overlap between targeted and non-targeted cell populations which makes it difficult to accurately enumerate cell populations. Light scattering from non- targeted white blood cells will be increased by the

greater size of cells to which only a few beads become attached in a reversible, sticky collision or by non¬ specific cellular adsorption. Similar size (about two micron diameter) beads coated only with adsorbed antibody do not show the extreme broadening along the light scatter axis of the histogram. Thus, a compact layer of antibody molecules adsorbed on the bead surface shows very little non-specific interaction with non-targeted cells. Storage of T4, T8-5X-Amdex-polystyrene beads at higher temperatures of about 20°C and 37°C for two weeks instead of at a refrigerated temperature of about 5°C further increased the broadening along the light scattering axis of the histogram. Unraveling of aminodextran chains or denaturation of conjugated antibody from its compact Y- or T-shaped structure to some uncoiled protein strand structure can further increase the preponderance of non¬ specific interactions with cells that lead to an increase in light scattering from non-targeted cell populations. Also, this denaturation of surface antibody does promote aggregation of the beads into large clusters which also show more intense light scattering.

It has been shown proteinaceous material, such as enzymes and antibodies, show increase thermal stability when such proteinaceous materials are immobilized by covalent attachment to solid surfaces or polymers. For example, urease on phospholipid-bound silica, (Kallury et al., Anal. Chem. , 64 : 1062-1068 (1992)) ; chymotrypsin on aldehyde/agarose gels, (Guisan et al . , Biotechnol. Bioencr. , 38 : 1144-1152 (1991)) ; glucoamylase on various periodate oxidized polysaccharides, Lenders and Crichton, Biotechnol. Bioeng. , 31: 267-277 (1988)) ; and K. Mosbach, Ed., Methods in Enzymology, Academic Press, New York, 1987, Vol. 135. Urease on a phospholipid-coated silica surface showed about 80-90% activity after being treated at 100°C for one hour or stored at 25°C for 168 hours.

Furthermore, U.S. Patent 5,120,6423, issued June 9, 1992 to Ching et al. , describes proteins that have been dried

onto solid substrates and resolubilized to provide active protein. Various proteins are used in immunoassay kits, containing large millimeter size plastic beads coated with an antibody, and distributed by Abbott Laboratories, North Chicago, Illinois.

In addition, sugars have been used in preservation in order to increase the stability of biological molecules and structures. For example, monosaccharides, disaccharides such as trehalose and sucrose, and other sugars have been found to be very effective in preservation of macromolecular structures (N. Hanafusa, Freezing and Drying of Microorganisms. Ed., T. Nei, University Park Press, Baltimore, 1969, pp. 117-129) as well as in protection of cell membranes during lyophilization and reconstitution (U.S. Patent 5,059,518 issued Oct. 22, 1991; and U.S. Patent 5,171, 661 issued Dec. 15, 1992) . Also, the naturally occurring polysaccharide, pectin, that is present in cell walls of all plant tissues and binds cells together, is widely used in the preservation of jellies and similar food products. There is evidence that some macromolecules such as triple helical soluble collagen show reversible reduction by about two-fold of layer thickness on glass as temperature is lowered from 15°C to 7.7°C (A. Silverberg, Polymer Adsorption and Dispersion Stability, Eds., E.D. Goddard and B. Vincent, ACS Symposium Ser. 240, Amer. Chem. Soc, Washington, D.C., 1984, pp. 161-167) .

The present invention improves prior systems by using two approaches to reduce the light scatter from the non- specific interaction between antibody-conjugated beads and cells and still retain a covalently-linked antibody-bead system. First, a short, low-molecular weight spacer, diaminoalkane is used as the coating for the beads to make the antibody-bead system structurally more closely resemble the adsorbed antibody-bead complex. Second, antibody-5X-aminodextran-coated beads were freeze-dried and reconstituted with distilled water to reduce non-

specific cell interactions with the polymeric sugar coating of the beads. Thus use of the present invention improves prior systems by decreasing the broadening along the light scatter axis, even at temperatures higher than that used in prior system. Freeze-drying and reconstitution of antibody-5X-aminodextran-coated beads further improves prior systems by increasing the stability of these beads.

DISCLOSURE OF INVENTION The present invention provides a method for reducing light scatter shift of non-targeted white blood cell population by using colloidal particles coated with diaminoalkane to make an antibody-bead system.

The present invention provides a method for reducing light scatter shift of non-targeted white blood cell population by freeze drying antibody-5X-aminodextran- coated beads.

The present invention also provides a method of increasing the stability antibodies when linked to 5X- aminodextran- and diaminoalkane-coated beads.

The present invention also provides a method of increasing the thermal stability of antibody-5X- aminodextran-coated beads by freeze drying said beads.

The present invention further provides a method for making antibody-diaminoalkane-coated beads.

The present invention provides antibody-colloidal particle systems for use on flow cytometry and VCS based instruments.

BRIEF DESCRIPTION OF DRAWINGS Figure 1 shows the reaction scheme for coating DAP on latex beads .

Figure 2 shows the binding curve (surface versus total antibody concentration) for CD4, CD8 antibody on DAP-coated latex beads.

Figure 3 shows the dependence of mean CD4/CD8 cell percent in whole blood lymphocytes on antibody (CD4, CD8) concentration in preparation of latex beads.

Figure 4 shows the binding curve for CD4, CD8 antibody on DAP-coated latex beads.

Figure 5 shows the binding curve for CD4, CD8 antibody on 5X-Amdex-coated latex beads.

Figure 6 shows the bead titer (μL of 1% w/v solids CD4, CD8-DAP-PS latex beads) dependence of mean CD4/CD8 cell percent in whole blood lymphocytes.

Figure 7 shows the histograms of lysed/quenched whole blood mixed with T4-DAP-PS beads stored at 5°C (a) , 20°C (b) , and 37°C (c) for two weeks. Same with T8-DAP-PS beads.

Figure 8 shows the histograms of lysed/quenched whole blood mixed with three preparations of T4-DAP-PS beads, pairwise presented for beads stored at 5°C and beads stored at 50°C (a-f) . Same with T8-DAP-PS beads (g-1) .

Figure 9 shows the histograms of lysed/quenched whole blood mixed with T4-5X-Amdex-PS beads (d-f) stored at 5°C (a) , 20°C (b) , and 37°C (c) for two weeks. Same with T8- 5X-Amdex-PS beads.

Figure 10 shows the histograms of lysed/quenched whole blood mixed with T4-5X-Amdex-PS beads stored at 5°C (a) ; freeze-dried and reconstituted and stored at 5°C (b) , 20°C (c) , and 37°C (d) for two weeks. Same with T8-5X- Amdex-PS beads (e-h) .

Figure 11 shows the histograms of lysed/quenched whole blood mixed with T4-5X-Amdex-PS beads freeze-dried

and reconstituted and stored at 5°C (a) and 50°C (b) for two weeks, and frozen and thawed (c) . Same with T8-5X- Amdex-PS beads (d-f) .

Figure 12 shows the pairwise histograms, before and after freeze-drying and reconstitution, of T4-5X-Amdex-PS beads, T8-5X-Amdex-PS beads, and 5X-Amdex-PS beads.

MODES FOR CARRYING OUT THE INVENTION

While the present invention is satisfied by embodiments in many different forms, there is shown in the drawings and will herein be described in detail a particular embodiment of the invention, with the understanding that the following description is to be considered as exemplary of the principles of the invention and is not intended to limit the scope of the invention as determined by the appended claims and their equivalent . The present invention provides a method for the preparation of colloidal particles with antibody covalently coupled to the bead surface, and for the verification of the suitability of these bead conjugates for use on flow cytometry and VCS based instruments such as the COULTER ^® STKS 2A and VCS instruments (Coulter Corporation, Miami, Florida) to enumerate white blood cell population in whole blood.

As used herein and in the claims, colloidal particles refer to polyacrolein beads, polyaldehyde beads and polystyrene beads, preferably polystyrene aldehyde/sulfate latex beads. The aldehyde/sulfate polystyrene latex particles used herein have reactive aldehyde groups which may combine with a substance such as an aminodextran and form covalent bonds. The size of the polymer particles ranges from about 0.2 to about 3.0 microns . The preferred particle size is in the range of about 0.6 to 3.0 microns, most preferred 1.5 to about 3.0. Because of the cell size dependence of shifts in the position of lymphocytes in forward versus side scatter and rf versus light scatter

histograms caused by surrounding cells of lymphocyte subsets with antibody-bead conjugates, the largest available beads of diameter about two microns are the most effective in giving a well-resolved shifted population of cells in the histograms.

As used herein and in the claims the term "alkane" in "diaminoalkane" refers to straight chain hydrocarbon of 2 to 6 carbon atoms, such as ethane, propane, butane, pentane hexane, preferably 3 and most preferably 1,3- diaminopropane (DAP) .

All reference to the monoclonal antibodies (mAb) used herein is by the identifying designations used by Coulter Corporation, Miami, Florida for monoclonal antibodies made by Coulter Corporation. The following information further identifies the antibodies used herein. The use of these monoclonal antibodies is by way of example only and is not to be understood as limiting the invention. The term "CD" refers to "Cluster Designation" adopted by the International Workshops on Human Leukocyte Differentiation Antigens. A.T.C.C. is the American Type Culture Collection, Rockville, Maryland.

Antibody CD Description or Reference

Til CD2 Derived from hybridization of mouse NS/1-AG4 cells with spleen cells of BALB/cJ mice immunized with T cell chronic lymphocytic leukemia cells.

T4 CD4 As Til, but immunized with peripheral human T lymphocytes .

T8 CD8 As Til, but immunized with human thymocytes .

1D3 U.S. Patent No. 4,931,395; A.T.C.C. Deposit No. HB 9445.

KC-48 U.S. Patent No. 4,865,971; A.T.C.C. Deposit No. HB 9584.

M02 CD14 R.F. Todd et al. , J. Immunol. ,

126.:1435-1442 (1981) . B4 CD19 L.M. Nadler et al. , J. Immunol. ,

131:2228 (1985) . Bl CD20 T. Stashenko et al. , J. Immunol.,

125:1678-1685 (1980) . NKH-1 CD56 J. D. Griffin et al. , J. Immunol

110:2947-2951 (1983) . MY4 CD14 J. D. Griffin et al . , J. Immunol

£8=932-941 (1981) . T3 CD3 P.C.L. Beverly and R.E. Callard (IgGl) Eur. J. Immunol, 11:329-334

(1981) .

Polystyrene aldehyde/sulfate latex beads have chemically reactive aldehyde functional groups on their surface, which are an intrinsic part of the polymer chains that make up the latex particle. These surface aldehydes are first chemically coupled to a short, low- molecular weight spacer such as 1, 3-diaminopropane or to a polymeric polyamine such as 5X-aminodextran which, in turn, are activated with a heterobifunctional crosslinker to covalently link antibody molecules to the bead surface. The resulting bead-antibody conjugates contain tightly bound antibody which is not released into the medium when suspended in bovine serum albumin (BSA) buffer solutions or in whole blood with anti-coagulants and red blood cell lysing and fixing agents.

The chemistry of the initial step of coupling diaminoalkane to the surface of polystyrene/sulfate latex beads is outlined in the chart in Figure 1. The reaction of the aldehyde moiety with amine groups to form covalent Schiff ' s base linkages proceeded best at an alkaline pH of about 10 or above. Thus, the beads which were suspended in distilled water were mixed with DAP to give a pH above 11. Schiff's base carbon-nitrogen bonds are stable if they originate from aromatic aldehydes . The surface

aldehyde on the polystyrene beads might be aromatic if it were attached to the styrene ring in the polymer. Since the method that was used to prepare the beads at Interfacial Dynamics Corporation (IDC) , Portland Oregon, is not divulged, the Schiff's bases must be reduced with alkaline sodium borohydride solution to yield a stable carbon-nitrogen bond. The latter reductant also plays an additional role in the reduction of the unreacted surface aldehyde groups to surface alcohol groups that cannot further react non-specifically with amine groups in antibodies or antigens expressed on cell surfaces.

In principle, the direct reaction of amine groups within the antibody with the surface aldehyde groups of the beads is also possible. In practice, however, this direct reaction gave very low surface antibody coverages and very poor test results on the Coulter VCS instrument. The negative results from this direct coupling of antibody to bead probably stem from several adverse factors. Steric hindrance can prevent the close approach of the amino groups on or near the surface of the antibody molecule to the surface aldehyde groups of the beads. Since antibody molecules can be denatured at the high pH of 10 needed for optimum amine-aldehyde coupling and by reaction with borohydride solutions, optimum stabilization of the Schiff's base formed by direct coupling could not be achieved. Therefore, the direct antibody coupling reaction was not further pursued.

The well-established heterobifunctional cross-linker, sulfosuccinimidyl 4- (N-maleimidomethyl)cyclohexane-1- carboxylate (sulfo-SMCC) , was used to activate DAP-coupled beads for conjugation to iminothiolane activated antibody. Because of the longer arms projecting from the surface of sulfo-SMCC activated DAP-beads, this conjugation process works well to yield high surface antibody coverages on the beads and good test results on the VCS. The conjugation depends on the reaction between maleimide groups of the activated beads and sulfhydryl groups (intrinsic or from

iminothiolane activation of amine groups) of the antibody.

Methods for the preparation of aminodextrans with various degrees of amine substitution and of polystyrene aldehyde/sulfate latex beads with a covalent, first layer of aminodextran to which antibodies were conjugated are described in pending patent application, Serial No. 07/961,157, filed October 15, 1992, which is a Continuation-in-Part of U.S. Patent 5,169,759, issued December 8, 1992.

Methods of the Invention

Primary reagents were optimized in the following sequence: 1. DAP concentration; 2. sulfo-SMCC concentration; 3. bead % w/v solids; 4. antibody concentration during conjugation. Initial runs of T4- beads were done with 1% w/v solids beads and concentrations of sulfo-SMCC and iminothiolane-activated antibody taken from established runs with previous beads, i.e., 6.75 μL of 10 mg/mL sulfo-SMCC per mL of 1% w/v solids beads and 0.625 mg I .T. -antibody per mL of 1% w/v solids beads.

1. DAP Concentration

DAP concentration was varied over a wide range. The stoichiometric amount of aldehyde on the bead surface was not analytically determined by the manufacturer; but, an upper limit on the amount of surface aldehyde was provided by knowing the amount of aldehyde that was used in the manufacture of the beads. This was 1 aldehyde per 50 A ² of surface area. For beads with a specific surface of 2.6464 m ² /g, this gives 0.088 μmol aldehyde per mL of 1% w/v solids beads. This requires 6.5 μg of diaminopropane per mL of 1% w/v solids beads by stoichiometry. In order to avoid aggregation of the beads, a much larger amount of DAP had to be used. In a series of runs with 2.0, 1.0, 0.5, 0.25, and 0.125% DAP with 1% w/v solids beads, the first trial with 2% DAP showed the least number and

smallest aggregates of beads after reduction with sodium borohydride and washing. Further trials with 2.0 to 16.0 mg/mL DAP were continued to conjugate T4 antibody and obtain VCS data. A DAP concentration of 4.0 mg/mL was sufficient to obtain reproducible %CD4 cells when 1% w/v solids beads were used. All higher DAP concentrations were also acceptable but the 4.0 mg/mL DAP concentration was selected for further work.

2. Sulfo-SMCC Concentration Next, several trials with 1/2X, IX, 2X and 3X the initial 6.75 μL of 10 mg/mL of sulfo-SMCC per mL of 1% w/v solids beads were made to activate the DAP-coated polystyrene aldehyde/sulfate beads. The process was carried through to conjugation of T4 antibody to the beads activated at different levels and testing of these beads in the VCS assay. The data show that 1/2X to 2X activation is acceptable but 3X activation is not acceptable for 1% w/v solids beads. Thus, the lowest activation level, 3.375 μL of 10 mg/Ml sulfo-SMCC per Ml of 1% w/v solids beads was used in further work.

3. Bead Concentration

Further trials were made with varying bead concentrations: 1%, 2.5%, and 4.2% w/v solids beads. Since the beads were supplied by the manufacturer at 4.2% w/v solids, this was used as the highest concentration of beads. In working with polystyrene aldehyde/sulfate beads of different mean diameters all at 1% w/v solids, it was found that beads of diameter 0.6 μm and smaller showed a much greater tendency to aggregate. Therefore, the particle number densities and mean free paths for these particle suspensions at 1% w/v solids were calculated and listed below:

Particles/cc Mean Free Path, μm 7 . 19 x 10 ¹¹ 1 . 12 8 . 29 x 10 ¹⁰ 2 . 29

Compared at a fixed percentage of solids, the much larger numbers of smaller particles and, thus, the smaller mean free path will produce many more collisions per unit time that can result in aggregation. Thus, 2 μm diameter beads at 4.2% w/v solids will have a mean free path of about 2 μm, so that aggregation might become a serious problem. The previously selected amounts of DAP and sulfo-SMCC for 1% w/v solids beads were multiplied by 2.5 and 4.2 for the respective trials, since surface area or number of aldehyde groups on the spherical bead surface is directly related to the mass of beads expressed for a constant volume through the specific surface in units of m ² /g. The VCS data for T4-conjugated beads at various bead concentrations showed that 2.5% and 4.2% w/v solids beads can be used throughout the preparative process a well as 1% w/v solids beads. The highest bead concentration, 4.2%, was used in further work to obtain the greatest yields of antibody-conjugated beads in any one trial.

4. Antibody Concentration

To optimize the amount of activated antibody, T4 or T8, used for conjugation to activated beads, a wide range of total 2-iminothiolane-antibody concentrations between 0.1 and 2.0 mg/mL were used. For each concentration trial, data were accumulated indirectly for surface antibody concentrations after the conjugation period (by filtering beads and reading absorbance at 280 nm for antibody concentration in supernatant; knowing total antibody concentration, the surface antibody concentration was calculated from the difference between total and supernatant antibody concentrations) and for %CD4 and %CD8 in whole blood by using the purified beads in the VCS assay. The results for 2 μm raw beads showed that surface coverages of T4 and T8 antibody on beads reached a plateau

at about 0.8 and 1.3 mg/mL total antibody concentration and 0.45 and 0.75 mg/mL surface antibody concentration, respectively; while %CD4 and %CD8 in whole blood by the VCS assay levelled off to about 0.3 and 0.8 mg/mL total antibody concentration, respectively. A graphical display of the latter results is shown in Figures 2 and 3. Similar data on a different lot of 2 μm raw beads gave surface coverage of T4 and T8 antibody with no well- defined plateau, especially the curve for T8 antibody which had a steadily climbing surface antibody load after the inflexion point. The results are shown graphically in Figure 4. The maximum surface coverage of T8 antibody on beads in both trials is about a factor of 5/3X greater than the maximum T4 antibody coverage. The above indirect method of assaying for surface antibody coverage does not discriminate between covalently-bound antibody and loosely-bound or adsorbed antibody, which may later be washed away by the bead purification steps with BSA buffer. For T4 antibody on beads at pH = 7.2-7.3 the maximum coverages in the 3-4 mg/m ² range represent about monolayer surface coverage when compared with literature data for IgG on solid polymer surfaces. Eosinisothiolcyanate-bovine serum albumin (EITC-BSA) on polymethyl methacrylate (PMMA) was shown by Tilton, Channing, and Gast [R.D. Tilton, Surface Diffusion and Hvdrophobic Interactions in Protein Adsorption, Ph.D. dissertation, Stanford University, 1991] to have a surface loading of about 1.5 mg/m ² at its plateau when concentrations were determined by a surface radioactivity technique. Similar data, about 1.5 mg/m ² saturation coverage, were obtained with radioiodinated BSA on sulfate polystyrene latex beads (Donald E. Brooks, Emerging Concepts in Medical Diagnostic Applications of Latexes. Orlando, Florida, December 4-5, 1989) . Bagchi and Birnbaum (Bagchi and Birnbaum, J. Colloid Interface Sci. , 83 : 460-478 (1981) ) have reported relatively low surface coverages, about 1.5 to 3.0 mg/m ² , of rabbit IgG on

negatively charged latex beads at pH = 4.0 and pH = 10.0; however, at the isoelectric point, pH = 7.8, of the IgG the saturation coverage was about 7.5 mg/m ² . The large variation in surface coverage with pH was attributed to changes in structure and/or orientation of the IgG, which in turn would give values of surface area per molecule between 3.1 x 10 ³ and 12.4 x 10 ³ A ² .

Therefore, the high surface coverages of about 5.5 to 7.0 mg/m ² of T8 antibody on beads at pH = 7.2-7.3 could represent either conformation/orientation changes in T8 antibody to accommodate more molecules per unit surface area or multi-layer adsorption of T8 antibody on top of the first covalently-bound T8 layer on the beads. BCA direct surface antibody assay results (vide infra) for samples taken after washing conjugated and blocked T8- beads with lxPBS showed that a large amount of loosely- adsorbed T8 was removed by the washings. An explanation for the second adsorbed layer of T8 antibody on beads but no similar layer of T4 antibody on beads lies in the isoelectric points of the two antibodies (pi range of T4 antibody: 5.4-5.8; pi range of T8 antibody: 6.7-7.3 from electrophoresis data) and the pH of the conjugation mixture in lxPBS, 7.3-7.3. Since the pi range of T8 and the pH of the medium during conjugation overlap considerably and antibody is least soluble or most attracted to itself at its isoelectric point, the excess T8 antibody tends to precipitate onto the solid bead surface on top of the first covalently-bound layer. T4 antibody on beads does not behave like T8 on beads since there is no overlap between its pi range and the pH of the medium during conjugation. Further, high coverages of both T4 and T8 on the same polystyrene aldehyde/sulfate beads, coated with polymeric 5X-aminodextran instead of the low-molecular weight, 1,3-diaminopropane, were observed in the 7.0-8.0 mg/m ² range. The surface coverage results, shown in Figure 5, indicate that polyamine-coated polystyrene latex particles are capable of increasing the

available surface for conjugation of antibody up to two¬ fold. Also, the aminodextran coating alters the surface charge of the particles from slightly negative due to terminal sulfate groups on polystyrene chains to positive due to protonated amino groups at pH 7.2-7.3 of the lxPBS medium. This positively charged particle surface is then more attractive to T4 antibody which has a slight negative charge at near neutral pH values of the medium. The T4 and T8 antibody concentrations during conjugation were therefore chosen to be 0.7 and 0.8 mg/mL for optimum surface coverage and VCS assay results.

Three trials of both T4 and T8 beads were done under the above optimized conditions. The volume of about 4% w/v solids beads was kept at 75 mL for each of the six runs, three with T4 antibody and three with T8 antibody. All trials were prepared with total T4 and T8 concentrations of about 6.2 and 7.1 mg/m ² (0.7 and 0.8 mg/mL) . The first trials gave apparent surface T4 and T8 coverages of 3.7 and 4.2 mg/m ² ; the second, 5.0 and 6.3 mg/m ² ; and the third, 5.3 and 6.2 mg/m ² , respectively.

Bicinchoninic acid (BCA) direct T4 and T8 antibody assay results for the second trial, in which samples were washed with lxPBS at pH = 7.2-7.3 immediately after conjugation, gave T4 and T8 surface coverages of 1.9 and 1.1 mg/m ² , respectively. These coverages are more in line with values expected for covalently-bound antibody on the latex beads.

All VCS testing was done with the following protocol: 20 μL of 1% w/v solids T4 or T8 beads and 100 μL of whole blood; two minute immediate mix. The standard statistical protocol for obtaining %CD4/CD8 data is two donors; eleven runs per donor with IX titer (20 μL of 1% w/v solids beads) ; and eleven runs per donor with 1/4X titer (5 μL of 1% w/v solids beads) . All three trials gave CVs under 5% for mean %CD4/CD8. A typical graphical representation of %CD4/CD8 from VCS assays versus bead titers with whole blood is shown in Figure 6 for trial

one. The curves all reach a plateau at about 10 μL titer of 1% w/v solids beads per 100 μL of whole blood aspirated on the VCS instrument, so that the standard titer of 20 μL of 1% w/v solids beads is more than enough for evaluating %CD4/CD8 in all potential donors.

THERMAL STABILITY OF T4,T8-CONJUGATED BEADS

The effect of storage of antibody-conjugated beads at higher than normal temperature (about 5°C) in a refrigerator was tested in closed vials at room temperature (about 20°C) and in an incubator at 37°C. The adequacy was tested in terms of number and strength of multiple covalent binding sites from antibody to the bead surface to prevent the antibody from changing its active conformation and orientation with respect to the bead surface when the beads are brought to elevated temperatures for extended periods of time.

In the VCS assay of beads with whole blood, the functional test of antibody-conjugated beads stored at elevated temperatures is performed for comparison with beads stored at 4-10°C. Their ability to shift relevant subpopulations of white blood cells in the rf versus LS histogram according to the specificity of the antibody is evaluated. Any denaturation of the surface-bound antibody may be reflected in non-specific interactions with non- targeted cell populations. This typically broadens the lymphocyte and granulocyte-monocyte distributions along the light scatter (LS) axis of the histogram and may yield a false higher percentage of targeted cells because of mixing of shifted and unshifted cell populations. Also, the normal titer of beads might not be sufficient to shift all targeted cells so that a false lower percentage will be measured. The supernatants of all tested bead suspensions were also analyzed for free antibody by an EIA assay to verify whether the antibody was strongly bound to the bead surface.

Testing was done on all three trials of optimized T4 and T8 beads, stored for two weeks in capped vials at 5°C in a refrigerator, at about 20°C ambient room temperature, and at 37°C in an incubator. VCS test results are presented in Table I and representative rf versus LS histograms are displayed in Figure 7. T4,T8-conjugated beads stored at elevated temperatures showed very little degradation as seen in the CVs ranging between 1.0 and 3.0% for the mean percent CD4 or CD8 for three temperatures. Individual coefficient of variations (CVs) for measurements made in quintuplet were high, 6.9 and 7.7%, in two runs with T8-conjugated beads stored at 37°C for two weeks; however, deviations from the mean in both cases were still less than two SDs. These T8-conjugated beads also showed the largest broadening of non-targeted lymphocyte and granulocyte-monocyte populations in the histograms shown in Figure 7. The degree of broadening can be associated with some degradation in the way in which surface antibody presents itself to cells with which it interacts, i.e., the extent of non-specific interactions with non-targeted cells has increased. The extent of deterioration, however, was not great enough for the population of non-targeted cells to interfere with the population of targeted cells in the VCS assay.

Table I

VCS Assay Results for %CD4/CD8 in Whole Blood

(Donor 32442 for lots -47 and -81;Donor 53254 for lot -24! for Accelerated Stability Run on T4, T8-Conjugated

Polystyrene Beads

Duplicate samples were also stored for two weeks at about 47-50°C in an oven. Representative VCS histograms for the heated samples and the control samples stored at about 5°C are shown in Figure 8. The T4 beads were still functional giving CVs of 3.3, 1.6, and 2.5 for mean percentages of CD4 of 53.09, 52.37, and 52.89, respectively, in the three trials, compared to control bead values of 52.73 (1.2% CV) , 52.52 (2.7% CV) , and 52.48 (2.2% CV) , even though some broadening of the non-targeted cell populations was observed with the use of heated beads. T8 beads stored at 47-50°C showed more extensive broadening of non-targeted cell distributions. They also failed completely to shift the targeted cells in one trial in which the control beads gave a mean %CD8 of 15.43 (2.4% CV) , and gave CVs of 4.4 and 7.6 for mean %CD8 of 16.20 and 15.07, respectively, in the other two trials, for which the control beads gave values of 15.62 (5.3% CV) and 15.20 (7.7% CV) .

EIA analyses for antibody in the supernatant of samples stored in closed vials at 20°C and 37°C showed about the same low amounts of free antibody as in analyses of refrigerated samples. Results for T4 samples gave the following data -- trial one: 1.13, 1.40, and 2.65; trial two: 0.75, 1.02, and 1.40; trial three: 1.43, 1.83, and 3.41 μg/mL at 5, 20, and 37°C storage, respectively. T8 samples gave the following results -- trial one: 0.58, 0.67, and 1.28; trial two: 0.44, 0.45, and 0.45; trial three: 1.24, 1.60, 2.64 μg/mL at 5, 20, and 37°C storage, respectively. Three samples stored at 37°C for two weeks showed larger than normal amounts, 2.65, 3.41, and 2.64 μg/mL, of free antibody. The largest amount represents about 1.7% of the surface-bound T4 antibody for freshly prepared beads.

Testing of T4,T8--5X-Amdex--polystyrene beads stored at 5°C, about 20°C, and 37°C for two weeks in closed vials showed considerable degradation of bead performance at the higher temperatures, especially for T4 beads, as seen in

the histograms in Figure 9. The percentage CD4/CD8 obtained with these beads stored at elevated temperatures could not be reliably measured. EIA analyses of the supernatants, however, showed little shedding of any antibody at elevated temperatures: for T4 beads- 0.43,

0.80, and 1.37 μg/mL; for T8 beads- 0.27, 0.57, and 0.55 μg/mL for 5, 20, and 37°C storage, respectively.

CRYOGENIC AND DRY STORAGE STABILITY OF T4,T8-CONJUGATED BEADS In one run, samples of T4,T8--DAP--bead trials one, two, and three, suspended in 1% BSA, 0.1% sodium azide in lxPBS, were freeze-dried and stored in capped vials in the dry state at about -20°C for four days. In a second run, duplicate samples as well as adsorbed T4 and T8 beads were freeze-dried and stored in capped vials in the dry state at about -20°C for one week. The lyophilized samples were then brought to room temperature and reconstituted with distilled water. These samples were all tested on the VCS for functionality and results are summarized in Tables II and III. In a third run, duplicate samples, first washed free of salt and protein with distilled water, were freeze-dried and stored in capped vials at -20°C for one week before reconstitution with BSA buffer solution. Test results for these beads compared to untreated samples are compiled in Table IV. VCS test results for run one freeze-dried beads showed excellent results for CD4 and CD8 percentages in the lymphocyte population, with CVs less than 5% except in one case in run two for T8 beads of trial three. Some small degree of expansion of the non- targeted lymphocyte population in the rf versus LS histogram did occur along the LS axis for bead trials two and three, while trial one beads showed hardly any change from results for non-freeze-dried bead suspensions. These results confirm the cryogenic and dry storage stability of covalently immobilized T4 and T8 antibodies on polystyrene beads. It is to be noted that the medium in which the

beads are freeze-dried is crucial for successful reconstitution of viable antibody-conjugated beads. When antibody-conjugated beads were first washed with distilled water as in the protocol for run three; then, VCS testing of the resultant beads showed much more deterioration in the functioning of T4 and T8 conjugated beads. Non- targeted lymphocyte and granulocyte-monocyte populations showed extreme broadening along the light scatter (LS) axis, similar to results obtained with T4 and T8 conjugated to aminodextran-coated polystyrene latex beads. T8 beads gave CVs greater than 5% for CD8 percentages.

Table II

VCS Test Results for Freeze-Dried and Reconstituted Samples (FD)

Table HI

VCS Test Results for Freeze-Dried, Dry Stored (1 Week), and Reconstituted Samples (FD)

Table IV

VCS Test Results for Distilled-Water- Washed Beads; Freeze-Dried and Reconstituted (FD)

Similar freeze-drying and dry storage of T4,T8--5X- aminodextran--polystyrene beads followed by reconstitution with distilled water showed a vast improvement over non- freeze-dried samples in performance in the VCS assay. Fresh and reconstituted sample histograms are compared in Figure 10. The freeze-dried and reconstituted, T4,T8-5X- Amdex-bead samples now perform as well as T4,T8--DAP- polystyrene or adsorbed T4,T8--polystyrene beads (CVs below 5%) and show histograms in which the L, G, and M distributions are as narrow along the LS axis. It is inferred that loose, hydrophilic chains of the polymeric sugar on the surface of the bead followed the water during its removal by freeze-drying to condense onto the particles. Then, the outer part of the aminodextran sugar layer collapsed or shrunk onto the bead without loss of conjugated antibody activity. Since aminodextran was

covalently coupled to the aldehyde bead surface and crosslinked with glutaraldehyde, as well as blocked with fresh aminodextran, a very low-density, polymer coating was fixed to the polystyrene beads. Thus, it is appropriate that the aminodextran coating and antibody conjugated to the outer surface of the coating might rearrange their structures upon freeze-drying. The antibody presentation on the composite bead surface apparently became the dominant surface feature so that non-specific interactions with non-targeted cells were enormously reduced.

Further, a sample of the reconstituted beads that were stored at about 4°C, was taken and stored in a closed vial in an incubator at 37°C for two weeks. These samples were also compared in the VCS assay and showed in Figure 10 a small amount of broadening of non-targeted cell populations for reconstituted beads kept at 37°C, but did not reverse to the broad distributions given by the original beads. Another set of freeze-dried and reconstituted T4,T8-5X-aminodextran bead samples, kept at about 47-50°C for two weeks, also passed specifications in the VCS assay as shown in histograms in Figure 11, showing that these beads were much more stable to thermal treatment than T4,T8-DAP-PS beads by not reversing to their original state. The mean percentage CD4 was 50.96 (1.8% CV) with control beads and 51.84 (3.1% CV) with heated beads, and the mean percentage CD8 was 15.82 (8.3% CV, < 2 SD) with control beads and 15.66 (7.3% CV, < 2 SD) with heated beads, all measured in quintuplet for a single blood donor. Thus, when freeze-dried and reconstituted, the polysaccharide derivative, aminodextran, coupled to polystyrene beads is much more effective in preserving the macromolecular structure of antibody that is conjugated to it than it is when not freeze-dried and reconstituted. Size measurements of the T4-5X-Amdex-PS beads before and after freeze-drying and reconstitution were made by photon correlation spectroscopy or quasi-elastic light

scattering on the COULTER N4MD sub-micron particle analyzer, using a 90° scattering angle. Beads suspended at 1% w/v solids in 1%BSA, 0.1% sodium azide in lxPBS buffer were diluted 300-fold with two times filtered (0.2 μm cellulose nitrate filter) lxPBS solution and measured in a 1 cm path quartz fluorescence cell with 5 sides polished. Reconstituted beads gave a weighted mean diameter of 3.04 μm with an intra-assay standard deviation of 0.34 μm while original beads gave similar data but with an additional large contribution (>15%) of dust, i. e., particles greater than 10 μm in size. The raw polystyrene aldehyde/sulfate beads gave a mean diameter of 2.07 μm with an intra-assay standard deviation of 0.36 μm and an inter-assay S.D. of 0.10 μm for three runs of eleven measurements each. The thickness of the T4 or T8 antibody and 5X-aminodextran coating is therefore 0.48 μm for spherical particles after freeze-drying and reconstitution. This compares with a calculated thickness of 0.22 μm for 5X-aminodextran on 0.30 μm diameter ferrite particles, as estimated from the carbon analysis for distilled water washed and dried particles. A few large aggregates of coated beads are responsible for large light scattering contribution, and these large bead aggregates are dissociated into individual, primary beads in the freeze-drying and reconstitution process.

The same T4,T8-5X-aminodextran-PS beads were also measured before and after freeze-drying and reconstitution on the COULTER Profile II flow cytometer to obtain forward light scatter versus side light scatter histograms as shown in Figure 12. A single predominant circular and tight distribution was observed for both coated and non- coated raw beads; however, both fresh and freeze-dried and reconstituted samples showed a minor weak distribution at higher forward and side scatter intensity showing the presence of small aggregates of two or three particles. No difference was observed in the histograms to indicate

the presence of very large aggregates in the original samples .

T4,T8--5X-Amdex--polystyrene beads that were frozen at -20°C and then allowed to thaw out in a refrigerator at about 5°C for about two hours were also tested on the VCS instrument. These samples showed a small amount of improvement in the histograms (Figure 11) by a narrowing of non-targeted L, G, and M distributions along the LS axis . All publications cited in this specification are indicative of the level of skill of those in the art to which this application pertains. Each publication is individually incorporated herein by reference in the location where it is cited. The following examples serve to illustrate the present invention. The concentration of reagents, temperatures and values of other variable parameters are only to exemplify application of the present invention and are not to be considered as limitations thereof.

EXAMPLE I

COUPLING OF 1, 3-DIAMINOPROPANE TO POLYSTYRENE LATEX BEADS

To 150 mL of 4.2% w/v solids polystyrene aldehyde/sulfate beads (Interfacial Dynamics Corp., about 2 μm diameter) in a 250 mL-polypropylene centrifuge tube were added 16.8 mg of liquid DAP (0.888 g/mL specific gravity) per mL of bead suspension or 2.838 mL DAP. The resulting suspension was mixed well by vortexing, sonicated for 30 sec, and roller mixed for 24 to 72 hours. Reduction of Schiff's base groups and unreacted aldehyde on bead surface.

Solid sodium borohydride, 47.46 mg per mL of 4.2% w/v solids beads that were roller mixed with DAP, or 7.119 g were added directly to the bead reaction mixture. No excessive effervescence took place in the presence of

excess DAP at a pH of about 11.5. The reaction mixture was roller mixed for three hours with occasional brief 30 sec sonication and then the beads were separated by centrifugation at about 2000 g for 5 min. The supernatant was discarded and the residue of beads was resuspended in 150 mL distilled water. Redispersion of the beads was accomplished by vortexing and brief sonication. The washing procedure by centrifugation was repeated four times and the final bead suspension was adjusted to 150 mL total volume with lxPBS solution.

EXAMPLE II ACTIVATION OF DAP COUPLED BEADS WITH SULFO-SMCC

About 14.175 μL of freshly prepared 10 mg/mL sulfo- SMCC in lxPBS was used per milliliter of 4.2% w/v solids DAP-coated polystyrene bead suspension. In a typical preparation, 1.063 mL of the sulfo-SMCC solution was added to 75 mL of 4.2% w/v solids beads. The mixture was then roller mixed in a 250 mL plastic centrifuge tube for about one hour, separated by centrifugation, and washed a plurality of times with lxPBS solution.

The functionalized, DAP-coated particles resulting from the above series of steps have pendent maleimidyl groups and are suitable for conjugation to a variety of biological molecules. If the substance which is desired to be conjugated to the particles has a sufficiency of active sulfhydryl groups, activation of that substance is not necessary, and the following Example III may be skipped.

EXAMPLE III ANTIBODY ACTIVATION WITH 2-IMINOTHIOLANE

HYDROCHLORIDE

A 35.99 mg/Ml concentrate of T4 monoclonal antibody in lxPBS containing 0.1% sodium azide was prepared. For

100 rag T4 (or T8) antibody and 15 mg/mL antibody concentration during coupling, the total reaction volume should be 6.667 mL. Using a 15:1:: IT:T4 activation ratio, 9.375 μmol (1.29 mg) IT (0.65 mL of 2 mg/mL IT) in lxPBS is required. Therefore, 3.238 mL of lxPBS solution was added to 2.779 mL of T4 concentrate, to which resulting solution an additional 0.65 mL of 2 mg/mL IT solution was added. The net resulting solution was roller mixed in a tube reactor for 1 hour. The contents of the reaction tube were then applied to the top of a 200 mL G-50

Sephadex column, equilibrated and washed with 500 mL lxPBS. The derivatized antibody was eluted using lxPBS and a plurality of 5 mL fractions were collected with the aid of a UV monitor. Fractions in the middle of the band absorbing at 280 nm were pooled and the A280 value was used to determine the T4/IT antibody concentration. Typically, the T4/IT or T8/IT concentration was about 5 mg/mL.

EXAMPLE IV CONJUGATION OF T4/IT OR T8/IT WITH

SULFO-SMCC DERIVATIZED PARTICLES

With a total volume 100 mL, the concentration of particles was 3.15% w/v solids and the T4/IT concentration was 0.7 mg/mL or the T8/IT concentration was 0.8 mg/mL. In one run, when the purified T4/IT solution concentration was 5.013 mg/mL, then 13.964 mL of T4/IT antibody solution in lxPBS was added to 75 mL of 4.2% w/v solids sulfo-SMCC activated particles which had been preconcentrated by the removal of 5 mL of supernatant. In another run, the purified T8/IT antibody solution concentration was 5.703 mg/mL, so that 14.028 mL of T8/IT antibody solution in lxPBS was added to 75 mL of 4.2% w/v solids sulfo-SMCC activated particles which had been preconcentrated by the removal of 5 mL of supernatant. The antibody/IT solution was added to the particles in 2 mL increments with

vortexing and ultrasonication between additions. The resultant mixture was then roller mixed in a 250 mL-tube for about two hours.

EXAMPLE V BLOCKING UNREACTED MALEIMIDYL AND SULFHYDRYL GROUPS

Unreacted maleimidyl groups on the sulfo-SMCC activated particles were blocked with L-cysteine after antibody conjugation. Typically, 11.88 mL of 5 mg/mL L- cysteine in lxPBS were added to the conjugation mixture of the previous step and the resulting suspension was roller mixed for 15 min. Unreacted sulfhydryl groups were blocked by the addition of 13.31 mL of 20 mg/mL iodoacetamide in lxPBS followed by the addition of 2.48 mL of 1M, pH 9.8 sodium borate buffer solution. The resulting suspension was roller mixed for 30 min, the blocked conjugation mixture was separated by centrifugation and the particles washed two times with a solution of 1% bovine serum albumin (Pentex Fraction V, protease free) and 0.1% sodium azide in lxPBS (BSA buffer solution) . After washing, the particles were resuspended in BSA buffer solution to a total volume of 75 mL ( 4.2% w/v solids) , roller mixed for 1 hour, stored at about 4°C for a time in the range of 8-16 hours, separated by centrifugation and washed three times with BSA buffer solution.

EXAMPLE VI COUPLING OF AMINODEXTRAN COATING TO BEADS

5X-aminodextran was prepared, purified, and freeze- dried according to methods described in a copending patent application, Serial No. 07/961,157, incorporated herein by reference. The same aldehyde/sulfate polystyrene latex particles of about 2 μm diameter from Interfacial Dynamics

Corporation (Portland, Oregon) were used here as for the preparation of DAP-coated beads.

59.524 L of polystyrene bead concentrate, 4.2% w/v solids, was added to a 250 mL centrifuge tube. 1.250 g of solid 5X-aminodextran were dissolved in 100 mL of distilled water and transferred into the tube containing the polystyrene particles. Then, 90.476 mL of distilled water was further added to the same tube to make a 1% w/v solids suspension of beads and a 5 mg/mL solution of 5X- aminodextran. Further, the suspension was made 1 mM in potassium hydroxide by the addition of 50 μL of 5M KOH solution to promote the reaction of aldehyde groups on the bead surface with amino groups of the aminodextran at a pH of about 10.0. The mixture in the tube was then roller mixed for a period of 8 to 16 hours.

The aminodextran coated beads prepared in the above manner were separated by centrifugation at about 2000 g for 10 min, the supernatant was discarded, and the residue was resuspended in a 2% w/v solution of 5X-aminodextran to make a total volume of 250 mL. The bead suspension was vortexed to redisperse the particles, made 1 mM in KOH by addition of 50 μL of 5M KOH solution, and then roller mixed for 1.5 hours. 2.925 mL of 25% glutaraldehyde solution (0.778 mmol or 1 mole per mole amino groups in 5X-aminodextran) were added to the bead suspension and roller mixed for one hour. The beads were then separated from free glutaraldehyde and aminodextran by centrifugation, the supernatant liquid discarded. To block unreacted aldehyde groups, the beads were resuspended in 250 mL of 1% w/v 5X-aminodextran and roller mixed for a time in the range of 8-16 hours. The beads were then separated by centrifugation, the supernatant was discarded, the residue was resuspended in 250 mL of a 10 mG/mL sodium borohydride in 1 mM KOH solution, and the bead mixture was roller mixed for another one hour. Subsequently, the beads were washed four times using centrifugation and lxPBS solution. The wash supernatant

liquids were discarded and the washed beads were suspended in sufficient lxPBS to yield 250 mL of 1% w/v solids 5X- aminodextran coated polystyrene latex beads.

The beads and antibody were activated and conjugated by procedures similar to those described above for T4 or T8 antibody and DAP-coated polystyrene beads except the bead concentration was 1% w/v solids, the sulfo-SMCC concentration was 4.2 times lower, and the quantity of T4/IT and T8/IT that was used during conjugation was 8 mg/m ² x 2 . 6464 m ² /g x 1 g/l00 mL or 0 . 212 mg/mL .

The invention has been described with reference to specific examples, materials and data. As one skilled in the art will appreciate, alternate means for using or preparing the various aspects of the invention may be available. Such alternate means are to be construed as included within the intent and spirit of the present invention as defined by the following claims.